The invention relates to a method for the operation of a centrifugal pump driven by an electric motor with variable frequency, wherein too small a flow through the pump is ascertained by monitoring electrical quantities.
Such a method is known from EP 0 696 842 A1. In that method, a standard frequency-voltage relationship is monitored in use. A current in the intermediate circuit is also monitored. When it is found that the value of the current is smaller than that which should be expected for the normal frequency-voltage ratio, it is assumed that the pump is operating without a load. In such a case the inverter is switched off and the motor stopped.
The electric motor of a pump of that kind is normally also cooled by the fluid being pumped. Consequently, protective measures have to be taken to prevent the pump from being destroyed when there is no through-flow. Such a situation may arise, for example, when the inflow pipe is blocked or when a valve therein has been closed in error. In such a case, the liquid remaining in place is heated, possibly to boiling point, and the pump or parts thereof and adjacent pipes can be destroyed as a result of the temperature or pressure surges.
Sensors in the pipes or reservoirs are often used to determine whether or not there is sufficient fluid present. Such sensors operate by optical means or are in the form of mechanical floats, but in all cases they are susceptible to malfunction and require a certain amount of maintenance.
In the known case, therefore, the current was used as an electrical quantity for the purpose of determining whether there exists a condition in which there is no through-flow. The control or monitoring fulfils its function, but only in a relatively narrowly circumscribed range of operation.
The problem underlying the invention is to detect, by simple means, when there is no through-flow present.
The problem is solved in a method of the kind described at the beginning by ascertaining the electrical power and comparing it to a control quantity formed as a function of the frequency of the motor.
This approach is no longer dependent upon a fixed threshold or limit value which, if it is not met, initiates a routine leading, finally, to the pump motor being stopped. Instead, the threshold value is modified dynamically in accordance with the operating frequency of the motor. By that means, it is possible to detect whether or not through-flow is present with significantly greater accuracy and irrespective of whether the motor is being operated at its nominal operating point or of whether its speed of rotation differs therefrom. The method is therefore especially suitable for centrifugal pumps that operate over a wide speed-of-rotation range, for example for the purpose of regulating the pumping rate, as is disclosed in DE 199 31 961 A1. The invention is based on the fact that the power consumption of a centrifugal pump decreases along with a decrease in the through-flow. When such characteristics are plotted with the motor frequency as a parameter in a power/through-flow diagram, a clear connection between through-flow and power is obtained in the region of relatively small amounts of through-flow.
The control quantity is preferably ascertained with the aid of a reference power that applies at a predetermined reference frequency. The predetermined reference frequency can be taken, for example, from the data sheet for the pump. The data sheet will normally showxe2x80x94for a specific reference frequencyxe2x80x94the power that has to be consumed in order to drive the pump even without any through-flow. If, however, the actual motor frequency differs from the reference frequency, it is not possible for the electrical motor power to be compared to a reference value directly. The reference power is therefore converted as a function of the actual frequency and the reference frequency so that the corresponding control quantity, which can be used for the comparison, can be obtained.
The control quantity preferably includes a product, one of the factors of which can be specified by a user. As a result, due account is taken of the fact that different users require different approaches to critical situations. Users having a higher safety requirement will select a factor that is correspondingly higher. In that case, a case of malfunction will be indicated, and/or a malfunction treatment routine will be initiated, together with stopping of the motor, even when there is still a small through-flow present. Other users who are more accepting of risk can approach the loading limit for the motor and then in fact stop the motor only when there is no longer any through-flow at all. Freedom of choice is provided by the simple means of using that factor.
Special preference is given therein to selection of a factor that is greater than unity. In that, it is assumed that the actual power basically cannot be less than the motor""s theoretically smallest power. Consequently, specifying that the control quantity is always formed using a factor that is greater than unity makes it possible always to remain on the safe side and rules out the possibility of errors by the user.
In an advantageous embodiment, at least two measurements of the power of the motor are made at different frequencies and without flow through the centrifugal pump, and a basis for the control quantity is ascertained therefrom. This approach is not dependent even on knowing the nominal output of the motor at nominal frequency. In contrast, however, it does becomes possible, with this approach, to take further losses into account, for example those that can occur in an inverter feeding the electric motor with variable frequency.
In this case, special preference is given to ascertaining the basis in accordance with the following formula:       G    fix    =                    G                  f          2                    -                                    G                          f              1              *                                ⁡                      (                                          f                2                                            f                1                                      )                          3                    1      -                        (                                    f              2                                      f              1                                )                3            
wherein Gfix: fixed power loss
f1: first frequency
f2: second frequency
Gf1: electrical power of the motor at frequency f1 
Gf2: electrical power of the motor at frequency f2.
This approach takes into account electrical power from effects which do not directly find expression in the delivery power of the pump. Determination of the control quantity becomes significantly more accurate using a power value of that kind.
The control quantity is preferably determined in accordance with the following relationship:       G    x    =            [                                    (                                          G                                  f                  1                                            -                              G                fix                                      )                    xc3x97                                    (                                                f                  x                                                  f                  1                                            )                        3                          +                  G          fix                    ]        xc3x97    F  
wherein fx: actual frequency
Gx: control quantity
F: factor
and the other quantities are as indicated above. It will be recognized that the control quantity is determined as a function of the frequency, with electrical powers (losses) not attributable directly to the delivery power of the pump additionally being taken into account.
The invention relates also to a pump arrangement having a centrifugal pump, an electric motor which drives the centrifugal pump, a controlled frequency converter which feeds the electric motor, a sensor device and an evaluating device.
In this pump arrangement the problem described above is solved by means of the fact that the sensor device ascertains values for determination of the electrical power, and the evaluating device has a dynamic limit value former, which forms a control quantity as a function of the frequency of the motor.
By means of a pump arrangement of this kind it is possible, by relatively simple means, to carry out monitoring of through-flow or absence of through-flow without having to accept major uncertainties if the motor operating frequency differs from a reference frequency.